Re: [Xen-devel] [PATCH 04/24] autoconf: trim the configure script; use autoheader

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Tue, 2012-04-17 at 14:18 +0100, Ian Jackson wrote:
> Ian Campbell writes ("Re: [Xen-devel] [PATCH 04/24] autoconf: trim the 
> configure script; use autoheader"):
> > On Mon, 2012-04-16 at 18:17 +0100, Ian Jackson wrote:
> > > Remove a lot of unnecessary tests.  Specifically, we no longer test
> > > for standard POSIX facilities which we expect to be provided
> > > everywhere and which we don't in any case have any alternative for.
> > 
> > A good rule of thumb might be that if it isn't provided by "apt-get
> > install build-essential" (or the common set of stuff from the equivalent
> > meta-packages across common distros) then configure should check for it
> > in the interest of providing a useful upfront error message?
> 
> Yes, I think so.
> 
> > > Switch to generating config.h.in with autoheader.
> > 
> > > @@ -132,7 +127,7 @@ AC_SUBST(libext2fs)
> > >  AC_CHECK_LIB([gcrypt], [gcry_md_hash_buffer], [libgcrypt="y"], 
> > > [libgcrypt="n"])
> > >  AC_SUBST(libgcrypt)
> > >  AX_CHECK_PTHREAD
> > > -AC_CHECK_LIB([rt], [clock_gettime])
> > 
> > -lrt is always available? -l<one-or-two-letters> libraries seem to be
> > the ones which tend to differ across platforms, despite being
> > standardised...
> 
> The current configure script throws away the result of this test, so
> it is definitely useless.  The effect in configure is to perhaps add
> -lrt to LIBS but we do not export configure's LIBS to every tools
> build.

It aborts if -lrt isn't available though, doesn't it?

> > > +AX_CHECK_PTYFUNCS
> > 
> > You don't actually add this until the next patch.
> 
> Oops.
> 
> Ian.



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel